【摘要】 为了解决局部通风机根据下一时刻需风量提前进行风速调整问题，提出一种基于改进海洋捕食者算法MMPA(Modified Marine Predators Algorithm)优化的Elman神经网络方法进行需风量预测。首先，利用混沌映射初始化种群以改善群体位置的不均匀性，引入反向学习对每次迭代前的个体进行反向操作，迭代中后期对猎物矩阵引入差分操作，并选择若干测试函数对其进行测试。其次，采用改进后的海洋捕食者算法优化Elman神经网络中的初始权值和阈值，提高需风量预测结果的精度。结果表明，MMPA-Elman神经网络模型预测精度更高，实现了风量的准确预测，为煤矿安全生产提供了保障。更多还原

【Abstract】 In order to solve the problem of local fans adjusting wind speed in advance according to the next time demand volume,a novel Elman neural network algorithm based on modified marine predator algorithm(MMPA)is proposed to predict the demand volume. Firstly,the chaotic map is used to initialize the population to improve the inhomogeneity of the population position,reverse learning is introduced to operate the individuals before each iteration,differential operation is introduced to the prey matrix in the middle and late iteration,and several test functions are selected to test it. Secondly,the improved marine predator algorithm is used to optimize the initial weights and thresholds in the Elman neural network to improve the accuracy of the required air volume prediction results. The results show that the prediction accuracy of MMPA-Elman neural network model is higher,and the accurate prediction of air volume is realized,which provides guarantee for the safety production of coal mines.更多还原